Change and persistence with failed technological innovation

Organizations learn by reinforcing past actions. When firms experience actions with negative outcomes, they change in response. Any persistence with such actions is said to be suboptimal, attributable to factors such as threat rigidities. This paper theorizes that persistence may sometimes be rational, attributable to the characteristics of feedback. It uses a novel dataset of medical devices to show how organizations learn to search for future novel or incremental innovations based on failure. It finds that firms change innovation activities when novel innovations fail, but persist when incremental innovations fail. These findings support the hypothesis that inferences based on failure from incremental innovations tend to be more robust, leading firms to be more willing to persist when failure occurs . Copyright © 2014 John Wiley & Sons, Ltd.